WE1.M1.5

Toward data quality evaluation in pre-train dataset for remote sensing foundation model

Rui Liu, Hongsheng Zhang, The University of Hong Kong, China; Jing Ling, Guangdong University of Technology, China
Session:
WE1.M1: The Prospect of Remote Sensing Foundation Models: From Generation to Application Oral
Track:
Community-Contributed Sessions
Location:
Room M1
Presentation Time:
Wednesday, 6 August, 09:00 - 09:15
Session Co-Chairs:
Hongsheng Zhang, The University of Hong Kong and Rui Liu, The University of Hong Kong
Presentation
Discussion
Resources
No resources available.
Session WE1.M1
WE1.M1.1: SAMPOLYBUILD: ADAPTING THE SEGMENT ANYTHING MODEL FOR POLYGONAL BUILDING EXTRACTION
Chenhao Wang, Jingbo Chen, Yu Meng, Yupeng Deng, Kai Li, Yunlong Kong, Aerospace Information Research Institute, Chinese Academy of Sciences, China
WE1.M1.2: SELF-SUPERVISED PRE-TRAINING FOR MULTI-TEMPORAL REMOTE SENSING IMAGERY BY JOINT EMBEDDING LEARNING
Qi Guo, Wuhan University, China; Jue Wang, Beijing Institute of Technology, China; Yinhe Liu, Yanfei Zhong, Wuhan University, China
WE1.M1.3: SSL-LIP: A two-stage Pre-training foundation model for SAR images
Yi Yang, Qingchen Fang, Xiaokun Zhang, Haipeng Wang, Fudan University, China
WE1.M1.4: MULTISPECTRAL TO HYPERSPECTRAL USING PRE-TRAINED FOUNDATIONAL MODEL
Ruben Gonzalez, Remote Sensing Technology Institute, German Aerospace Center (DLR), Germany, Germany; Devyani Lambhate, Joao Lucas de Sousa Almeida, Paolo Fraccaro, Benedikt Blumenstiel, IBM Research, India; Conrad Albrecht, Remote Sensing Technology Institute, German Aerospace Center (DLR), Germany, Germany; Thomas Brunschwiler, IBM Research, Germany; Nassim Ait Ali Braham, Remote Sensing Technology Institute, German Aerospace Center (DLR), Germany, Germany; Ranjini Bangalore, IBM Research, India
WE1.M1.5: Toward data quality evaluation in pre-train dataset for remote sensing foundation model
Rui Liu, Hongsheng Zhang, The University of Hong Kong, China; Jing Ling, Guangdong University of Technology, China
Resources
No resources available.